Candidate List Generation

1. A method comprising: receiving, by one or more processors, a query image including an object; receiving, by the one or more processors, a plurality of indexed images that match the object; computing, by the one or more processors, a region of interest for the object; computing, by the one or more processors, an overlap between a first region of interest corresponding to a first indexed image and a second region of interest corresponding to a second indexed image; determining, by the one or more processors, that the overlap between the first region of interest and the second region of interest satisfies a threshold; and including, by the one or more processors, the first indexed image and the second indexed image in a candidate list of indexed images.

2. The computer-implemented method of claim 1 , further comprising: computing an area match for each of the plurality of indexed images; and sorting the plurality of indexed images to generate a list of recognition results based on the area match.

3. The computer-implemented method of claim 2 , wherein the list of recognition results includes the first indexed image and the second indexed image.

4. The computer-implemented method of claim 2 , wherein the area match for an indexed image is a ratio of a convex hull of matching inlier points included in the indexed image and a corresponding region of interest.

5. The computer-implemented method of claim 2 , further comprising: computing an overlap between the region of interest corresponding to the first indexed image and a region of interest corresponding to a next remaining indexed image in the list of recognition results; determining that the overlap between the region of interest corresponding to the first indexed image and the region of interest corresponding to the next remaining indexed image in the list of recognition results satisfies the threshold; and including the next remaining indexed image in the candidate list of indexed images.

6. The computer-implemented method of claim 1 , wherein the region of interest for the object is a polygon denoting a boundary of the object in the query image.

7. The computer-implemented method of claim 1 , further comprising: providing the candidate list of indexed images for display to a user.

8. A system comprising: one or more processors; and a memory storing instructions, which when executed cause the one or more processors to: receive a query image including an object; receive a plurality of indexed images that match the object; compute a region of interest for the object; compute an overlap between a first region of interest corresponding to a first indexed image and a second region of interest corresponding to a second indexed image; determine that the overlap between the first region of interest and the second region of interest satisfies a threshold; and include the first indexed image and the second indexed image in a candidate list of indexed images.

9. The system of claim 8 , wherein the instructions cause the one or more processors to: compute an area match for each of the plurality of indexed images; and sort the plurality of indexed images to generate a list of recognition results based on the area match.

10. The system of claim 9 , wherein the list of recognition results includes the first indexed image and the second indexed image.

11. The system of claim 9 , wherein the area match for an indexed image is a ratio of a convex hull of matching inlier points included in the indexed image and a corresponding region of interest.

12. The system of claim 9 , wherein the instructions cause the one or more processors to: compute an overlap between the region of interest corresponding to the first indexed image and a region of interest corresponding to a next remaining indexed image in the list of recognition results; determine that the overlap between the region of interest corresponding to the first indexed image and the region of interest corresponding to the next remaining indexed image in the list of recognition results satisfies the threshold; and include the next remaining indexed image in the candidate list of indexed images.

13. The system of claim 8 , wherein the region of interest for the object is a polygon denoting a boundary of the object in the query image.

14. The system of claim 8 , wherein the instructions cause the one or more processors to: provide the candidate list of indexed images for display to a user.

15. A non-transitory computer-readable medium storing instructions which, when executed by one or more processors, causes the one or more processors to: receive a query image including an object; receive a plurality of indexed images that match the object; compute a region of interest for the object; compute an overlap between a first region of interest corresponding to a first indexed image and a second region of interest corresponding to a second indexed image; determine that the overlap between the first region of interest and the second region of interest satisfies a threshold; and include the first indexed image and the second indexed image in a candidate list of indexed images.

16. The non-transitory computer-readable medium of claim 15 , wherein the instructions cause the one or more processors to: compute an area match for each of the plurality of indexed images; and sort the plurality of indexed images to generate a list of recognition results based on the area match.

17. The non-transitory computer-readable medium of claim 16 , wherein the list of recognition results includes the first indexed image and the second indexed image.

18. The non-transitory computer-readable medium of claim 16 , wherein the area match for an indexed image is a ratio of a convex hull of matching inlier points included in the indexed image and a corresponding region of interest.

19. The non-transitory computer-readable medium of claim 16 , wherein the instructions cause the one or more processors to: compute an overlap between the region of interest corresponding to the first indexed image and a region of interest corresponding to a next remaining indexed image in the list of recognition results; determine that the overlap between the region of interest corresponding to the first indexed image and the region of interest corresponding to the next remaining indexed image in the list of recognition results satisfies the threshold; and include the next remaining indexed image in the candidate list of indexed images.

20. The non-transitory computer-readable medium of claim 15 , wherein the region of interest for the object is a polygon denoting a boundary of the object in the query image.